React + TypeScript + Vite

🏠 Livably

AI-powered platform helping college students create personalized roommate agreements

Live URL: https://roomate-contract.web.app/ | [Video Walkthrough]

🎯 Problem

The majority of college roommate conflicts stem from unclear expectations about shared living spaces. Students need a structured way to align on preferences before conflicts arise.

💡 Solution

Livably guides household members through a preference assessment, then uses AI to generate a customized living agreement covering quiet hours, cleanliness, guests, and more.

✨ Key Features

• 🤝 Multi-user household management with shareable invite links
• 📝 12-question living preference assessment
• 🤖 Gemini AI-powered contract generation
• 📄 Downloadable PDF agreements
• ☁️ Real-time sync across household members

🛠️ Tech Stack

• Frontend: React, TypeScript, Vite, Tailwind CSS
• Backend: Firebase (Firestore, Hosting, Authentication)
• AI: Google Gemini API
• CI/CD: GitHub Actions

🏗️ Architecture Highlights

• Designed normalized Firestore schema for households, users, and contracts
• Implemented multi-step form with progress tracking and validation
• Built prompt engineering system for consistent AI contract generation
• Created responsive UI supporting mobile and desktop experiences

🚀 Getting Started

[Installation instructions]

🎓 What I Learned

• Integrating LLM APIs into production workflows
• Managing multi-user state with Firebase
• Designing user flows for collaborative experiences
• Balancing AI creativity with structured output requirements